Thread by @SDullien

Important contribution to the output gap debate: Marcell Göttert & Timo Wollmershäuser from @ifo_institut present an approach how to use survey data to estimate output gaps. Some short comments in a #thread 1/

econpol.eu/publications/w…

Göttert and Wollmershäuser use survey data on capacity utilization to estimate output gaps. In comparison to the current approach by the EU_Commission, this can already be seen as an improvement as the measure is less susceptible to revisions of past values. 2/

However, the new measure has still some weaknesses: 1. While the path of output gaps from surveys (that is the ups and downs) are plausible, the methodology does not offer any theory-based indication when it is negative and when it is positive. 3/

The authors use the long-term average (over 20 years) and implicitely assume that when capacity utilisation is at the long-term average, the output gap is closed. The problem with this approach is that a longer period of 4/

negative output gaps is ruled out by assumption. If, for example because of macroeconomic policy mistakes, an economy remains below full capacity for a number of years, this slowly becomes the new normal. 5/

While this possibility is regularly assumed away in macroeconomic models/analysis, especially for member states of the euro-area which do not have their own central bank anymore, it is plausible that a negative output gap can persist for several years. 6/

Hence, especially for cases like Italy, it is plausible that the presented measure is misaligned concerning the closed output gap level. 7/

Moreover, an extreme event like the #Covid19 crisis (which shaved more than 10 % of Italian GDP in 2020) would still change the output gap by about 0,5 percentage points (if a 20-year-period is used for computing the average). 8/

Problem 2: The measure seems more to reflect the microeconomic capacities than the macroeconomic capacity. With (logically) only existing firms included in the survey, a wave of insolvencies 9/

which would push up unemployment could at the same time close the output gap. Hence, the measure neglects that underutilized labour in the form of unemployment can be moved back into firms. 10/

Any level of uneployment could in principle be compatible with a closed output gap according to this methodology, even if capacity extensions are easily possible without significant price pressure. 11/

So, what to make of it? Maybe combine the use of surveys with the approach by @RobinBrooksIIF and @heimbergecon who try to include measures of inflationary pressure into output gap estimates? 12/

CC @achimTruger @sprofit1 @jakob_eu @clausMichelsen @PhilippaSigl @pm_steinberg @hogrefeJens @tom_krebs_ @katjaRietzler @CR_Proano #CANOO /END